The Internet of Things (IoT) is often imagined as a vast web of communicating objects – from cars, to houses, to wearable devices, even coffee makers. But as users get caught up in the conceptual hype of IoT, they often forget that its purpose is to generate meaningful data – lots of it. As Loraine Lawson explains in an article for IT Business Edge, the deluge of data associated with IoT may cause a flood of problems for CIOs. Here she addresses five infrastructure / data analytics solutions to keep technology chiefs high and dry.
Five-Alive IoT Questions
- What do you mean by real-time data?
- How will you adapt data analytics to the demands of IoT?
- What are your storage needs for IoT data?
- What are your network needs for the IoT?
- What’s your API strategy for IoT data?
The first tip to keep in mind is that real-time isn’t always…well there’s really no such thing as real time. Each embedded object will respond in its own time, the time determined by the developer based on the urgency of the application. Emergency situations require a millisecond response, whereas simply monitoring may allow for considerable lag time (whole seconds). LTE or WiFi are not the same as Bluetooth or 2G.
Then there is the data itself. Like a pipe leak, sensor data aggregates into puddles, but soon becomes pond-sized until you’re suddenly facing an ocean of information requiring “real”-time analytics – a heck of a lot of code for processing. The only real solutions are to either run processing such that only exceptions are sent to network or send it to the cloud.
Beyond speed and analytics, the IoT generates new storage needs, and data managers need be particularly choosy about what gets stored and for how long. Some use cases require immediate response, while many more cases yield a trove of historical data that must be stored for analysis. It’s also important to consider the effects of app-to-app interaction, rather than human-to-app interaction – especially when dealing with a network of applications. Build a system with thousands of sensors and you’re going to have issues with inbound bandwidth.
Lastly, it’s no longer a world of closed doors and anchored employees. IoT technology is another boundary breaker, and for your workforce to enjoy its benefits, there may be a greater need for APIs both internally and externally. All those sensors and embedded apps need to count for something, and that something is useful information at the fingertips of the user.
Read the original article at: http://www.itbusinessedge.com/blogs/integration/five-questions-to-help-cios-avoid-iot-data-problems.html